Safe anytime-valid inference: from theory to implementation in psychiatry research

The work was supported by the Dutch Research Council (NWO), as part of the Enabling Personalized Interventions (EPI) project in the Commit2Data–Data2Person program under contract 628.011.028. The work for this thesis was carried out at CWI and UMC Utrecht.

Chapters 2 until 7 are based on work previously published as stand-alone papers:

Chapter 2, Generic E-Variables for Exact Sequential k-Sample Tests that allow for Optional Stopping, has been published in the Journal of Statistical Planning & Inference, doi url: https://doi.org/10.1016/j.jspi.2023.106116

Chapter 3, Exact Anytime-valid Confidence Intervals for Contingency Tables and Beyond, has been published in Statistics and Probability Letters, doi url: https://doi.org/10.1016/j.spl.2023.109835

Chapter 4, Information extraction from free text for aiding transdiagnostic psychiatry: constructing NLP pipelines tailored to clinicians’ needs, has been published in BMC psychiatry, doi: 10.1186/s12888-022-04058-z

Chapter 5, Bayesian network analysis of antidepressant treatment trajectories, has been published in Scientific Reports, doi: 10.1038/s41598-023-35508-7

Chapter 6, Outcome Prediction of Electroconvulsive Therapy for Depression using a Bayesian Network Model based on Clinical Information, has been published in Psychiatry Research, doi: 10.1016/j.psychres.2023.115328

Chapter 7, Safe Sequential Testing and Effect Estimation in Stratified Count Data, has been published in the PMLR conference proceedings, PMLR 206:4880-4893 and can be accessed through https://proceedings.mlr.press/v206/